HypoArena: leaderboard
Metric: Average of the six domain ratings: Bradley-Terry-Davidson rating (base 1500, open-ended) from reference-independent pairwise judgments by seed-2.0-pro, both presentation orders, with the source-derived reference hypothesis set as an anonymous competitor; baseline single-pass generation; higher is better. Source: arxiv.org. Saturation forecast: Around September 2027. 15 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Sonnet 4.6 (High) | 1654 |
| 2 | Claude Opus 4.6 (High) | 1652.7 |
| 3 | GPT-5.4 (High) | 1631.4 |
| 4 | Kimi K2.6 | 1582.7 |
| 5 | GLM-5.1 | 1581.9 |
| 6 | DeepSeek V4 Pro (High) | 1535.4 |
| 7 | MiniMax-M2.7 | 1498.9 |
| 8 | DeepSeek V4 Flash (High) | 1484.6 |
| 9 | GLM-5 | 1464.3 |
| 10 | MiniMax-M2.5 | 1423.8 |
| 11 | GPT-5.4 Mini (High) | 1421.9 |
| 12 | Gemini 3.1 Pro (Preview) (High) | 1394 |
| 13 | Gemini 3 Flash (High) | 1345.7 |
| 14 | Kimi K2.5 (Thinking) | 1289.5 |
Interactive version: theaggregate.ai/benchmark?slug=hypoarena · How It Works · Data refreshed daily, snapshot 2026-09-29.